Use of smartphone apps to monitor human exposure to solar radiation: Comparison between predicted and measured UV index values

The Ultra-Violet Index (UVI) is widespread used to communicate the UV radiation intensity to the general public. The knowledge of the UVI value and of its daily variation is essential for many techniques for monitoring the personal exposure to UV radiation. The UVI values are usually provided by the meteorological services and nowadays it is very common to find UVI forecast tools even in smartphone apps. In this paper, with the aim to evaluate the prediction accuracy of six smartphone apps, a measurement campaign of UVI has been carried out. The measurements have been conducted for the site of Pisa (central Italy), using a portable photoradiometer equipped with a UV erythemal irradiance probe (operating range 250–400 nm). The measured UVI values have been compared with the predicted UVI values (using smartphone apps). Bland-Altman and Passing-Bablok methods have been used to compare the data, the comparison has been conducted on the basis of 90 different UVI measurements, taken at different times, different days and different sky conditions. From the comparison between measured and predicted UVI values it has been possible to observe a general poor accuracy of the apps. The percentage deviations between measured and predicted UVI values were quite high and only one app was able to predict more than 70% of the measured data with an average percentage deviation lower than 30%.